Grenoble France Paris France Flower Menus A New
Grenoble France Paris France Flower Menus: A New Type of Marking Menus with Large Menu Breadth, Within Groups and Efficient Expert Mode Memorization Gilles Bailly Eric Lecolinet Laurence Nigay 1
Introduction • Marking menus [Kurtenbach & al. 91] Novice Mode Expert Mode 2
Marking menus • Advantages – Circular design – Fluid Transition – Eyes-free Selection • Scale Independence Novice Mode Expert Mode 3
Marking menus • Limitations • Limited number of commands – Menu depth: number of levels (hierarchical Marking menus) – Menu breadth: number of commands at each level 4
Depth: multi-level Marking menus • Limitation – Error rate in expert mode – Real estate requirement Novice Mode Expert Mode 5
Multi-Stroke menus [Zhao 04] • Several simple marks instead of one complex mark – Temporal instead of spatial composition • Less error prone • Marks can be overlapped – Less amount of screen space Novice Mode Expert Mode 6
Marking menus • Limitations – Limited number of commands – Menu depth: number of levels (hierarchical Marking menus) – Menu breadth: number of commands at each level 7
Problem Number of commands at each level (menu breadth) 8
Menu breadth • Menu breadth Limited to 8 -12 commands Context menu in Power. Point 16 commands 9
Polygon menus Polygon Menus [Zhao & al. CHI’ 06] 16 commands 10
Polygon menus Polygon Menus [Zhao & al. 06] 16 commands 11
Polygon menus • Tangential gestures Radial Gestures • 3 actions – 1 clic – 1 shift – 1 stroke • Specific layout Polygon Menus [Zhao & al. 06] Impact on learning and memorization 12
Outline • Introduction • Flower Menus • Experiment • Conclusion 13
Flower menus • Key features – Large number of commands – Within groups (groupings) – Memorization 14
Flower menus 15
Large number of commands • Straight and curved gestures – 4 degrees of curvature 16
Degree of curvature Save as File Search Speech Pigtail Straight Bent Cusped 17
Pilot study • Users could draw gestures precisely enough ? • Answer: YES • Bias – Higher variability on diagonals • specific gesture recognizer – Takes into account user characteristics – Recognition rate: 99% for common configurations All Bent gestures for the counterclosewise 18
Large number of commands • 8 orientations • 4 degrees of curvature • 2 rotating directions ( Clockwise, counterclockwise ) For each menu level: • Width potential: 56 commands • But of course do not need that many 19
Distribution of the number of commands 14 12 10 8 6 4 2 0 1 -4 5 -8 9 -12 13 -16 17 -20 21 -24 25 -28 Marking menus Polygon menus Flower menus 6 common applications: Word, Excel, Adobe Reader, Firefox, Thunderbird, Photoshop 20
Flower menus • Large number of commands • Within groups (Groupings) • Memorization 21
Within Groups (groupings) • Semantic groups at the same level – Separated by a line – No title • hierarchical groups • Improve visual search The « File » Menu in Microsoft Word 5 within groups 22
Within Groups (groupings) One within group by branch One within group between two separators 23
Distribution of the number of commands in within groups 80 Most within groups contain 1 or 2 commands 70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Flower menus 6 common applications: Word, Excel, adobe reader, firefox, thunderbird, photoshop 24
Flower menus • Large number of commands • Within groups (groupings) • Memorization 25
Memorization • Improve learning and memorization – Highlight semantic relationships • Within groups – Visual organization • Proximity • Closure (same branch) 26
Flower menus • Large number of commands • Within groups (groupings) • Memorization • Eyes-free selection • Hierarchical 27
Multi-level Flower Menus • Flower menus work in the same way as Multi-Stroke menus – Several simple marks – Overlapped marks • 2 -level Flower menus – Potential: 56 * 56 > 3000 A 3 level Flower menus 28
Outline • Introduction • Flower Menus • Experiment • Conclusion 29
Experiment • Objective – Learning performance of the expert mode of: • Flower menus • Polygon menus • Linear menus (Base line) 16 commands in 7 within groups 30
Experiment • Intentional learning – Users asked to learn the expert mode of the 16 items • Procedure – Familiarization – Training phase: from novice to expert mode • Duration: 5 mn • Approximatively 2 -3 selections for each command – Testing phase: expert mode selections • Stimulus : the name of the item 31
Summary 18 participants X 3 menu techniques X 16 commands (16 gestures in expert mode) X 2 repetitions = 1728 selections We measured how many items the users are able to select in expert mode. - Number of items that are recalled 32
Results 81% 40% 35% Flower two times better than Polygon Surprisingly, performance of Linear close to Polygon 33
Possible explanations • Radial vs « tangential » gestures • In Polygon menus close items can have very different gestures • User comments – « difficult to recall the gesture while I know the position of the item » 34
Possible explanations • Marking menus efficiency – Learn the expert mode by repeating the same gesture in novice mode. • Polygon v. s. Flower – Both based on command repetition – But quite different results • A simple expert mode • A simple mapping between commands and gestures 35
Subjective Preferences • 17/18 chose Flower menus as their favorite technique • Best score for criteria: – – – Familiarization Simplicity Learning Speed Accuracy • User comments – They prefer « gestures starting from the center » with Flower menus rather than « having to perform two operations » with Polygon menus 36
Outline • Introduction • Flower Menus • Experiment • Conclusion 37
Conclusion • Flower menus – – – Curved gestures added Support a large number of commands (20) Support within & hierarchical groups Favor learning and memorisation of commands Eyes-free selection, Scale Independence, Fluid Transition • Learning experiment – Flower menus offer better performance than Polygon and Linear menus – We gave some possible explanations 38
Future Work • Improve gesture recognizer • Study Implicit Learning • Compare with classical marking menus – Breadth Organization vs Depth Organization – Impact on learning performance 39
u o y k n a h T Merci e i z a r G Video: www. gillesbailly. fr 40
- Slides: 40